An exhaust gas discharged from diesel engines contains particulate matter (PM) comprising as main components carbonaceous soot and soluble organic fractions (SOFs) comprising high-boiling-point hydrocarbon components, which are likely to adversely affect humans and environment when discharged into the air. Accordingly, ceramic honeycomb filters have conventionally been attached to exhaust pipes of diesel engines for removing PM. One example of ceramic honeycomb filters for capturing PM in the exhaust gas is shown in FIGS. 1 and 2. A ceramic honeycomb filter 10 comprises a ceramic honeycomb structure comprising porous cell walls 2 defining a large number of outlet-side-sealed flow paths 3 and inlet-side-sealed flow paths 4 and a peripheral wall 1, and upstream-side plugs 6a and downstream-side plugs 6c sealing the exhaust-gas-inlet-side end surface 8 and exhaust-gas-outlet-side end surface 9 of the outlet-side-sealed flow paths 3 and the inlet-side-sealed flow paths 4 alternately in a checkerboard pattern.
As shown in FIG. 2, this ceramic honeycomb filter 10 is gripped by support members 14 and longitudinally sandwiched by support members 13a, 13b in a metal container 12. The support members 14 are generally formed by metal meshes and/or ceramic mats. The ceramic honeycomb filter 10 mounted to a diesel engine receives mechanical vibration and shock from the engine, road surfaces, etc. via the support members 13a, 13b and 14. Because such large ceramic honeycomb filters as having outer diameters of more than 200 mm are subject to a large load by vibration and shock, they are required to have high strength.
Among the characteristics required for ceramic honeycomb filters, PM-capturing efficiency, pressure loss, and a PM-capturable time period (a time period until pressure loss reaches a predetermined level from the start of capturing) are important. Particularly the capturing efficiency and the pressure loss are in a contradictory relation, higher capturing efficiency leading to larger pressure loss and thus a shorter PM-capturable time period. A low-pressure-loss design provides low capturing efficiency, despite a long PM-capturable time period. To satisfy all of these contradictory filter characteristics, investigation has conventionally been conducted to provide technologies of controlling the porosity, pore size distribution, etc. of the ceramic honeycomb structure.
JP 2003-534229 A discloses a ceramic structure having a cordierite phase as a main component, and a thermal expansion coefficient of more than 4×10−7/° C. and less than 13×10−7/° C. between 25° C. and 800° C., its permeability and pore size distribution meeting the formula of 2.108×(permeability)+18.511×(total pore volume)+0.1863×(percentage of pores of 4-40 μm to total pore volume)>24.6.
JP 2007-525612 A discloses a filter for capturing diesel particulate matter, which has a median diameter d50 of less than 25 μm, and a pore size distribution and porosity meeting the relation of Pm≦3.75, wherein Pm is expressed by Pm=10.2474 [1/((d50)2 (% porosity/100))]+0.0366183 (d90)−0.00040119 (d90)2+0.468815 (100/% porosity)2+0.0297715 (d50)+1.61639 (d50−d10)/d50, wherein d10, d50 and d90 (d10<d50<d90) represents pore diameters at cumulative pore sizes (by volume) of 10%, 50% and 90%, respectively.
The technologies described in JP 2003-534229 A and JP 2007-525612 A restrict only pore structures (size and distribution) measured by mercury porosimetry to predetermined ranges, but they fail to design ceramic honeycomb filters capable of efficiently capturing nano-sized PM, which are considered to have particularly large influence on humans, with small pressure loss.
JP 2006-095352 A discloses a honeycomb filter having porosity of 45-70%, which has cell walls formed by a porous substrate having an average pore diameter A (μm) measured by mercury porosimetry, and an average pore diameter B (μm) measured by a bubble point method, an average pore diameter difference ratio [(A−B)/B]×100 being 35% or less, the average pore diameter B being 15-30 μm, and the maximum pore diameter measured by a bubble point method being 150 μm or less.
JP 2006-095352 A describes that the average pore diameter A measured by mercury porosimetry is a value reflecting the average diameter of pores on cell wall surfaces, while the average pore diameter B measured by a bubble point method is a value reflecting the minimum pore diameter in the cell walls, that therefore, in the case of cell walls having a pore structure as shown in FIG. 4 (a), in which pores in the cell walls have small diameters, while those on cell wall surfaces have large diameters, the average pore diameter B measured by the bubble point method is much smaller than the average pore diameter A measured by mercury porosimetry, and that on the other hand, in the case of cell walls having a pore structure as shown in FIG. 4 (b), in which pores in and on the cell walls have the same diameters, and in the case of cell walls having a pore structure as shown in FIG. 4 (c), in which pores in the cell walls are larger than those on cell wall surfaces, the average pore diameters A and B measured by mercury porosimetry and the bubble point method are not substantially different.
JP 2006-095352 A describes that cell walls having the average pore diameter difference ratio of 35% or less, namely small difference between the average pore diameter A measured by mercury porosimetry and the average pore diameter B measured by the bubble point method, have a structure which comprises a smaller number of large pores on cell wall surfaces than that of small pores in the cell walls [FIG. 4 (a)], a ratio of diameters in the cell walls to those on cell wall surfaces being relatively small; namely, there are many pores having similar diameters in and on the cell walls [FIG. 4 (b)], and many smaller pores on cell wall surfaces than those in the cell walls [FIG. 4 (c)]. Namely, the honeycomb filter described in JP 2006-095352 A is constituted by cell walls having many pores as shown in FIGS. 4 (b) and 4 (c).
Because honeycomb filters shown in Examples of JP 2006-095352 A have the maximum pore diameter measured by the bubble point method in a range of 129-145 μm, it is expected that pores in the cell walls have larger diameters. Accordingly, the honeycomb filters have insufficient efficiency of capturing nano-sized PM having particularly large influence on humans, despite small pressure loss.